copy the linklink copied!Chapter 2. Building scenarios for the long-run evolution of urban areas
This chapter presents the reference scenario, as well as the various counterfactual scenarios examined in the study. The first section details the projected evolution of exogenous factors such as population and income, which impact the results. The second section describes the reference scenario, in which all currently implemented policies are kept fixed and all announced policies materialise. The third section elaborates on the various counterfactual scenarios, which introduce new land use and transport policies in the period from 2019 to 2050.
A scenario is composed of two distinct elements. The first element is a set of hypotheses about the variables that evolve over time and play an important role in determining the outcomes of the modelling exercise, but are not affected by these outcomes. Such key variables, which are referred to as exogenous, include the various components of vehicles’ user costs, the urban population and the pre-tax prices of vehicles, fuel and electricity. The complete list of exogenous variables and their associated definitions is presented in Table 2.1.
The second set of inputs in a scenario consists of assumptions about the state of policy instruments that impact the outcomes of the modelling exercise. Important variables include: (i) policies determining average emissions per vehicle kilometre, such as the fiscal treatment of conventional and electric vehicles; (ii) policy instruments affecting the relative attractiveness of public transport alternatives (vis-à-vis that of privately owned vehicles), such as road pricing, on-street parking fees and public transport subsidies; and (iii) land use policy instruments that determine the driving distances and the degree of car dependency in the long run, such as density regulations.
The study generates outcomes that depend on the values assumed for the exogenous variables and policy instruments. These outcomes, which are referred to as endogenous variables, include, but are not limited to, the car ownership rate, the adoption rate of electric vehicles and the population density. Other key outcomes include the endogenous prices of land and housing, which vary across residential zones. As they are endogenous, the above outcomes cannot be used as inputs in the construction of scenarios. Instead, these outcomes are used to qualitatively and quantitatively evaluate the results of the modelling exercise. For instance, the adoption rate of electric vehicles in a given urban area is the outcome of certain policies, relevant prices and unique characteristics of the area. Thus, no valid scenarios can be built by assuming an ex-ante EV adoption rate. Rather, economic and environmental implications are calculated based on the EV adoption rate resulting from the exogenous and policy instrument variables defined under a given scenario.
The economic and environmental outcomes under different policy scenarios are evaluated against a single reference scenario. The reference scenario reflects a continuation of current land-use and transport policies into future years. It therefore constitutes a business-as-usual benchmark, against which other counterfactual policy options can be evaluated. Examples of counterfactual policies are tax and fee exemptions for electric vehicles or policies that lower the pecuniary cost of public transport relative to the benchmark level used in the reference scenario.
This chapter is organised in three sections. The first presents the evolution of the set of assumptions common to all scenarios. The second lays out the components of the reference scenario, in which all currently-implemented policies are kept fixed. The third describes the counterfactual policy packages, which introduce new policies to be evaluated.
copy the linklink copied!2.1. The root: common assumptions across all scenarios
The evolution of the exogenous variables presented in Table 2.1 is common across all scenarios. The model simulation particularly depends on two of these variables. These are the population growth rate and the speed at which the key attributes of electric vehicles, such as their battery price, lifetime and their driving range, improve over time.
The rate of urban population growth is a key determinant of model outcomes because it affects the aggregate demand for transport. A larger population translates to more trips of all types. Furthermore, population growth may also imply an increase in the per capita demand for transport, especially in monocentric urban settings where the points of interest such as jobs, shopping malls and other services are concentrated within a single central business district (CBD). In such cases, population growth tends to increase the distances between residential areas and the aforementioned points. Finally, an increase in population is likely to generate more traffic congestion and reduce driving speeds, which may increase gasoline consumption and therefore the amount of emissions per kilometre driven.
The evolution of the various attributes of EVs are also an important determinant of model outcomes insofar as they influence the adoption rate of these vehicles. A higher adoption rate of EVs results in a lower carbon footprint per kilometre driven by an average vehicle. The pecuniary attributes of EVs include their private cost components, such as the cost of battery depreciation and electricity consumption per kilometre. The non-pecuniary attributes of EVs include speed, which is similar to that of conventional vehicles, and driving range, i.e. the maximum distance that can be driven before the battery has to be recharged. The driving range determines the degree to which the vehicle could be utilised without the limitations imposed by the need for frequent recharging. A limited driving range is more restrictive when the spatial density of EV recharging stations is low.
The values for population growth are obtained as weighted averages of plausible boundary values. The upper bound annual population growth is set at 2.2%, a number that is based on the recent rate of population growth in Auckland (Auckland Council, 2018[1]). That rate implies a doubling of the population in the 32-year time horizon of the study (2018-2050). On the other hand, the lower bound annual population growth rate is 1.3% and constitutes a continuation of the long-run trends in the population growth of the city. That rate implies a population increase of approximately 50% in the time range of the study. The reference scenario weights these values equally, implying an annual population growth rate of 1.75%.
In line with the use of boundary values in population growth, the analysis uses a weighted average of an upper and a lower bound in the evolution of the relevant EV attributes. These bounds capture the uncertainty that characterises the rate of technological change in the electric vehicle industry. In the upper bound, i.e. the optimistic case, the average battery depreciation cost per kilometre falls from NZD 0.048 in 2018, to NZD 0.020 in 2030 and to NZD 0.016 in 2050. Simultaneously, the benchmark driving range of 200 kilometres in 2018 increases to 380 kilometres in 2030 and to 500 kilometres in 2050 (Laffont and Peirano, 2013[2]; Chediak, 2017[3]). The upper bound hence represents a continuation of the rapid decline in battery costs observed during in recent years (Nykvist and Nilsson, 2015[4]).
In the lower bound, i.e. the pessimistic case, technological progress reaches a plateau and then slows down. Due to that deceleration, the value of battery depreciation cost per kilometre falls only slightly, to NZD 0.043 in 2030 and NZD 0.039 in 2050. Battery driving range reaches 235 kilometres in 2030 and 250 kilometres in 2050. Reaching a technological plateau before the interim time point of the study (2030) implies that the cost of battery depreciation per kilometre remains at levels that may be high enough to prevent parity with ICE vehicles (Laffont and Peirano, 2013[2]).
The reference values are selected using weighted averages of the upper and lower bound values. The upper bound value is weighted at 66% and the lower bound is weighted at 34%. This reflects the tendency of historic EV cost projections to underestimate future progress in the industry (Laffont and Peirano, 2013[2]). The lower and upper bound values for population growth and technological change in the EV industry are summarised in Table 2.2.
All scenarios examined in the study contain a series of common assumptions. These common assumptions are hereafter referred to as the “root” of all scenarios.
In all scenarios, public open spaces of an urban area remain intact throughout the considered period. These spaces include open recreational facilities, parks and conservation areas. The same holds for areas that host: industries, retailers and office spaces; infrastructure such as roads and highways; educational and health facilities; as well as areas that serve special purposes (e.g. cemeteries, defence areas).
Furthermore, key individual preference parameters remain fixed. For instance, strong preferences for low-density development at year 2018 are assumed to persist until the terminal period, in 2050.1 However, the model allows housing consumption to adjust upwards as income grows. That adjustment is larger in the case of low-density housing types, reflecting the general preference for low-density development. A similar logic holds for preferences for leisure time.
Unlike preferences, which are time-invariant, other unobserved factors that evolve over time can play an important role in the choice of vehicle type. Such factors may include habits, beliefs, as well as practical and informational constraints that contribute to the current low penetration rate of EVs. One example of such a constraint is the status quo bias, which may stem from, for example, a lack of information of the current operational costs and attributes of EVs or an unwillingness to switch to a new product that has a different charging time and thus implies changing transport habits. The study assumes that the strength of such unobserved factors subsides over time. As shown in detail in Chapter 5, the asymmetry of these factors between ICE vehicles and EVs subsides slowly but steadily over time. All else equal, the 55% of the unobserved advantages ICE vehicles possess over EVs subsides is eliminated by 2050.
The fuel economy of conventional vehicles in Auckland, a key factor in the choice of vehicle type, is predicted to increase, from the observed average of 11.2 km per litre of gasoline in 2018, to 14.4 km per litre in 2030 and 17.4 km per litre in 2050.2 The projection covers both conventional (gasoline and diesel) and hybrid cars. It is based on a clear trend observed in the Auckland vehicle fleet between the years 2000 and 2018. The projection is constructed using separate trends in a series of vehicle attributes, such as body weight, engine size and body type. These factors are known to affect fuel consumption (EPA, 2018[5]; New Zealand Transport Agency, 2018[6]). Subsequently, these trends were extrapolated to the study period, using techniques that are described further in Chapter 4.
The fuel economy projections for conventional vehicles are used to calculate the fuel consumption and, by extension, CO2e emissions per kilometre. The latter is computed by multiplying the CO2e content of a litre of gasoline with the fuel economy (litres/km). That multiplication produces the CO2e emission per kilometre travelled, a value that falls over time as fuel economy improves. From a baseline value of 0.211 kgCO2e per km in 2017, it declines to 0.160 kgCO2e per km in 2030 and to 0.132 kgCO2e per km in 2050. This constitutes a projected 37% decline in emissions per kilometre between 2018 and 2050, as displayed in Figure 2.3.
The projections for the changes to the CO2e emissions per kilometre made in this study can be compared to projections made by the New Zealand Ministry of Transport (NZ MoT) in the Transport Outlook (New Zealand Ministry of Transport, 2017[7]). In the conservative “Base Case” the NZ MoT Transport Outlook projects a 22% decline in the CO2e emissions per kilometre of gasoline (petrol) vehicles for the period 2014/15-2039/40. There are two primary reasons why the carbon intensity of gasoline vehicles is projected to fall more in this study than in the Transport Outlook: first, the study makes projections to 2050 whereas the Transport Outlook projects changed to 2040. Extending the time period implies greater improvements in fuel economy. Second, the projection in this study covers both conventional vehicles and hybrid vehicles whereas the Transport Outlook projection covers only gasoline-powered vehicles. Hybrid vehicles are, on average, significantly more fuel-efficient than gasoline vehicles and are expected to increase as a share of the fleet. Their inclusion thus increases the expected improvements in fuel efficiency. The implications of the fuel economy projection of conventional vehicles for the results is fully explored in Chapter 6.
The energy efficiency of electric vehicles is projected using a linear extrapolation of the evolution of the energy consumption (kWh/km) of Nissan Leaf, the most popular EV model in New Zealand (New Zealand Transport Agency, 2018[6]; EVDB, 2019[10]). Over time, technological improvements will cause the electricity consumption per kilometre driven to fall. The underlying greenhouse gas emissions of EVs (CO2e/km) depend on their energy efficiency (kWh/km), as well as on the amount of carbon used in the generation of each energy unit (CO2e/kWh). Emissions per kilometre fall over time as the electricity required to drive one kilometre falls and the share of renewables in the New Zealand Electricity grid increases (Transpower, 2018[8]; IEA, 2017[9]). These projections are presented in Figure 2.2: the electricity consumption per kilometre falls, from 0.17 kWh per km in 2017 to 0.14 kWh per km in 2030 and to 0.09 kWh per km in 2050. Moreover, the emission factor of the grid falls from 0.119 kgCO2e per kWh in 2017, to 0.057 kgCO2e per kWh in 2030 and to 0.022 kgCO2e per kWh in 2050. This translates to a fall in the emissions per kilometre over time, a change that is shown in Figure 2.3: from a baseline of 0.020 kgCO2e per km driven for an electric vehicle in 2017 to 0.008 kgCO2e per km in 2030 and to 0.002 kgCO2e per km in 2050.
Public transport is projected to become fully electric by 2050, in alignment with the roadmap developed by Auckland Transport (Auckland Transport, 2018[11]). From 2025, Auckland Transport will only procure buses with zero tailpipe emissions and by 2040, the fleet is expected to be fully electric. In line with this procurement plan, the share of electric buses is projected to increase from 0% in 2018 to 25% in 2030 and 100% in 2050. The fuel efficiency of electric buses is projected to improve at the same rate as the one projected for electric passenger vehicles: from the benchmark value of 1.07 kWh per km in 2018, to 0.88 kWh per km in 2030 and to 0.57 kWh in 2050.
The carbon intensity of electric bus travel per passenger-kilometre is calculated by taking the product of three elements: the occupancy rate of buses, their fuel efficiency and the carbon intensity per kWh.3 The average occupancy rate per kilometre is estimated at an average of 7 passengers in a 35 passenger capacity bus.4 The occupancy rate is exogenous and is kept fixed across the study period.5 For example, under a policy scenario promoting public transport the occupancy rate of buses is likely to increase, leading to a fall in the CO2 emissions per kilometre. However, the potential error introduced will likely be insignificant in assessing total emissions in 2050, as electric buses will consume electricity from an almost clean grid. The study makes an assumption about the evolution of real income of a representative individual. All scenarios assume that this income will continue growing in line with the rates observed during the last three decades.
To extrapolate income growth, the study uses the historical evolution of the per capita disposable income in New Zealand, which in Figure 2.4 is expressed as a percentage deviation from its 2017 level. The linear trend, derived for the period 1990-2017, is used to generate positive deviations from the same base year throughout the study period. This implies a real disposable income that in 2030 is roughly 20% higher than that of 2017 and 54% higher in 2050.
A similar approach is taken to extrapolate electricity prices from the period of observation (2005-2018) to the study period (2018-2050). Real pre-tax residential electricity prices display a clear upward trend between 1980 and 2013, but stabilise in the period 2013-2018 (IEA, 2018[13]). The projected path, shown in the upper panel of Figure 2.5, shows pre-tax electricity prices increasing over time at a pace that gradually slows. This reflects potential technological improvements in power generation and efficiency gains in the use of electricity amid continued demand growth.
The common set of assumptions used in all scenarios is summarised in Table 2.3.6
copy the linklink copied!2.2. The reference policy scenario
The reference policy scenario is composed of the root assumptions detailed in the previous section and a set of currently- implemented land-use and transport policies, which are kept fixed throughout the study period 2018-2050. Hence, this reference scenario represents a “business-as-usual” (hereafter, BAU) path. The policy components of this scenario are presented in Table 2.4.
In transportation, a BAU path implies that the contribution of the Emission Trading Scheme (ETS) in the final price of gasoline remains fixed at the average value of 0.016 New Zealand dollars per litre of gasoline observed between the inauguration of the scheme in July 2010 and the base year of the study (2018). The same holds for all other tax components that make up the final price of gasoline such as the excise tax. Furthermore, a regional fuel tax, also a component of the final price of gasoline, is assumed to remain constant at NZD 0.115 per litre until 2050.
A BAU path also implies a fixed vehicle ownership tax that is not differentiated between conventional and electric vehicles. That tax is composed of three different charges: a registration fee (NZD 202.81 in 2018), a value-added tax paid on the purchase of a new vehicle (estimated at NZD 4 682 in 20187) and an annual licensing fee. The first two components are converted to flat kilometre taxes, resulting in a rate of NZD 0.0244/km8. In contrast, the annual licensing fee is paid every year the car remains active in the fleet. It is therefore modelled as a fixed yearly cost.
Furthermore, under the BAU path, the current public transport fares in Auckland, as well as the road user fees that apply countrywide (NZD 0.06 per kilometre in 2018), remain constant. The density of electric vehicle recharging stations, estimated to be 0.0343 stations per km2 of urban fabric in 2018, increases to 0.07 stations in 2030 and to 0.1 stations in 2050. Finally, the BAU path does not involve any ban of conventional vehicles.
In land use, a BAU path implies that all current urban development regulations remain fixed. The expansion of the current residential urban footprint, which is represented by the grey-coloured areas in Figure 2.6, takes place through the conversion of land that in year 2018 is labelled as “future urban”. The latter is represented by the black-coloured areas in Figure 2.6 and is conceded for development in a pattern that mimics the existing one. This means that, in the BAU path, the proportion of land allocated to each development type (e.g. single-family detached, single-family attached and multifamily) in the newly developed areas is the average share of residential land occupied by each of these types in the existing residential footprint. Finally, in a BAU path all infrastructure, such as roads, schools and healthcare facilities, expands proportionately to existing infrastructure.
The reference scenario, which is comprised of the “root” assumptions and the BAU policy components, constitutes the benchmark against which counterfactual scenarios are evaluated. The components of this reference scenario are summarised in Table 2.4. All counterfactual scenarios include the same assumptions about population growth and technological progress in the EV industry as in the reference scenario (Table 2.2). However, in every counterfactual scenario the business-as-usual (BAU) policy scenario is replaced by an alternative policy scenario. The latter is a combination of one or more policy packages, which are presented in the following section.
copy the linklink copied!2.3. Counterfactual policy packages
The selection of the counterfactual policy packages is motivated by the target of reducing emissions from urban transport. The selected policies lower emissions through different channels. Depending on the combination of the policy and the mechanism of emissions reductions, the shift towards low-carbon mobility may also generate various side benefits. For example, policies that reduce the number of private vehicles on the roads also reduce congestion, while policies that increase population density around employment zones can reduce travel distances (Ang and Marchal, 2013[16]).9 Another criterion for the policies selected is also that they are compatible with the land-use and transport model, MOLES, used to run the policy simulations.
The first package entails policy components designed to promote public transport over private vehicles (hereafter referred to as the “promote public transport” policy package). Policies in this package increase the fixed private costs of both ICE vehicles and EVs. This is done by increasing the annual circulation fees of these vehicles by NZD 2 000. The package also includes a substantial increase in the operational costs of private vehicles. Road charges are increased by NZD 0.5 per kilometre while a congestion charge scheme in the form of a double cordon toll surrounding the CBD and the isthmus area. The pricing rates are aligned with European examples, with daily crossing cost placed at 1.5% of the average gross daily income.10 As an offsetting measure, all public transport fares are given a permanent discount of 80%, compared to their 2018 levels.11 Finally, the package imposes a significant increase in the ETS tax component throughout the period from 2018 to 2050. This increases to NZD 1.16 per litre in both 2030 and 2050. The policy package comprising measures to promote public transport is presented in Table 2.5.
The second package is designed to promote electric vehicles over conventional private vehicles (hereafter referred to as the “promote electric vehicles” policy package). It includes a subsidy to EVs of NZD 2 000 in both 2030 and 2050. That subsidy represents both the monetary benefits of a direct purchase subsidy, e.g. a VAT reimbursement and annual circulation fee exemption, and all the indirect benefits that EVs may enjoy. Such indirect benefits may include the right to use bus lanes, free parking and other advantages that are not modelled explicitly. Moreover, the electric vehicle package alters the relative operating costs of EVs relative to ICE vehicles. The package exempts EVs from a steep increase of NZD 0.5 in the kilometre tax, which is imposed on ICE vehicles. Therefore, it generates a substantial difference in the operational costs of the two types of vehicles. The regional fuel tax component remains fixed at its benchmark level, i.e. at NZD 0.07 per litre. The “promote electric vehicles” policy package is summarised in Table 2.6.
In addition to the two policy packages described above, which comprise transport policies, the study also evaluates the effect of two sets of land-use policies. The first land-use policy package introduces a generalised densification in the entire urban area. The second policy package is a set of measures to densify a particular part of the Auckland urban area. There are four variations of the targeted densification policy package, which are displayed in the four panels of Figure 2.7.
The widespread and targeted densification policy packages are both implemented by relaxing vertical density (i.e. building height restrictions) and horizontal density regulations (see for a definition of vertical and horizontal density). This applies to all development types considered in the study (see Chapter 3 for a discussion of development types). Under the widespread densification package, all single-family (attached or detached) and multi-family dwellings are allowed to be 50% taller in 2030 and 2050. Simultaneously, the minimum undeveloped area, i.e. the percentage of the land plot surface occupied by the building is multiplied by a factor of 1.5 in all areas. For instance, if 30% of a land plot’s surface could be covered by buildings in 2018, widespread densification increases that rate to 45%. The study refers to that rate, which is an indicator of horizontal structural density, as the coverage coefficient.
In the widespread densification package, the horizontal and vertical densification occurs without altering the development typology across space: each land plot continues hosting the same development type as in 2018, but that development type is characterised by higher floor-to-area ratio and covers a larger fraction of the land plot. Finally, future urban areas are conceded entirely for multi-family housing rather than for all housing types and according to the existing proportions, as it is the case in the reference scenario. The measures included in all densification packages are summarised in Table 2.7. Widespread densification provides a crude way to increase the supply of residential floor space. Despite increasing building height and coverage coefficients uniformly across space, the package increases residential floor space by a larger amount in central areas. This occurs because the proportional adjustment of coefficients produces more floor space in dense development types and most of these types are more frequent in central areas.12 Therefore, widespread densification may also function as a way to concentrate residential space around key points of economic activity, such as jobs and large shopping hubs. That may help reduce vehicle kilometres and, by extension, the greenhouse gas emissions from private vehicles.
In order to examine the degree to which densification may help reduce emissions in urban areas, the study examines an alternative urban planning strategy, in which densification occurs exclusively in targeted areas. The study examines four targeted densification packages, each of them selecting the zones to be densified with a different criterion. The selected zones under each of these four packages are shown in Figure 2.7. The first targeted densification package, here referred to as “transit-oriented densification” and coded as TD1, selects areas of substantial density that lie close to large employment hubs and transit nodes. The second one, referred to as “CBD-surrounding densification” and coded as TD2, is a package that densifies low density areas surrounding the central business district. The third package, entitled “densify isthmus” and coded as TD3, selects all low-density areas in the entire inner core of Auckland as areas where densification may take place. Finally, the package entitled “job-surrounding densification” and coded as TD4, further densifies areas of already-considerable density surounding the largest employment hubs. Targeted densification packages intensify the growth of structural density, and therefore housing supply, in the selected zones. The associated building heights and coverage coefficients are provided in detail in Chapter 3.
Interaction between different policies and policy types is an important feature of the study. Transport and land-use policies are not necessarily additive in their outcomes. Rather, they are subject to synergies and trade-offs. This necessitates an examination of different combinations of the policy packages presented in this chapter. That is provided in the results of the study, detailed in Chapter 5.
References
[16] Ang, G. and V. Marchal (2013), “Mobilising Private Investment in Sustainable Transport: The Case of Land-Based Passenger Transport Infrastructure”, OECD Environment Working Papers, No. 56, OECD Publishing, Paris, https://dx.doi.org/10.1787/5k46hjm8jpmv-en.
[15] Auckland Council (2018), Auckland Plan 2050, http://www.aucklandplan.govt.nz.
[1] Auckland Council (2018), Auckland Plan 2050 Evidence report: Demographic trends for Auckland: Data sources and findings, Auckland Council, Auckland, https://www.aucklandcouncil.govt.nz/plans-projects-policies-reports-bylaws/our-plans-strategies/auckland-plan/about-the-auckland-plan/Evidence%20reports%20documents/evidence-report-demographics.pdf.
[11] Auckland Transport (2018), Auckland’s Low Emission Bus Roadmap.
[3] Chediak, M. (2017), The Latest Bull Case for Electric Cars: the Cheapest Batteries Ever, Bloomberg, https://www.bloomberg.com/news/articles/2017-12-05/latest-bull-case-for-electric-cars-the-cheapest-batteries-ever.
[5] EPA (2018), Fuel Economy Data, https://www.fueleconomy.gov/feg/download.shtml.
[10] EVDB (2019), Nissan Leaf price and specifications, Electric Vehicle Database, https://ev-database.org/car/1106/Nissan-Leaf.
[13] IEA (2018), Energy prices in national currency per unit.
[14] IEA (2018), Indices of energy prices by sector.
[9] IEA (2017), Energy Policies of IEA Countries - New Zealand, International Energy Agency, http://www.iea.org/t&c/.
[2] Laffont, K. and E. Peirano (2013), Annex to D3.1 - Technology Assessment Report: Demand-side Technologies: Electric Vehicles, Technofi, http://www.e-highway2050.eu/fileadmin/documents/Results/D3/report_electric_vehicles.pdf.
[7] New Zealand Ministry of Transport (2017), Transport Outlook: Future State.
[6] New Zealand Transport Agency (2018), New Zealand vehicle fleet open data sets, https://www.nzta.govt.nz/resources/new-zealand-motor-vehicle-register-statistics/new-zealand-vehicle-fleet-open-data-sets/.
[4] Nykvist, B. and M. Nilsson (2015), “Rapidly falling costs of battery packs for electric vehicles”, Nature Climate Change, https://doi.org/10.1038/nclimate2564.
[12] Statistics New Zealand (2017), Disposable income per person, http://archive.stats.govt.nz/browse_for_stats/snapshots-of-nz/nz-social-indicators/Home/Standard%20of%20living/disp-income-pp.aspx.
[8] Transpower (2018), Te Mauri Hiko - Energy Futures, https://www.transpower.co.nz/sites/default/files/publications/resources/TP Energy Futures - Te Mauri Hiko 21 May%2718 - web.pdf.
Two different concepts of density are used in this study: structural density and population density.
Structural density refers to the distribution of dwellings, building and residential floor space across the urban area. Two different measures of structural density are used: vertical density and horizontal density. Vertical density refers to building height and is measured through the floor-to-area ratio. The floor-to-area ratio is the relationship of the total amount of floor space to the parcel of land the building is located on. It is calculated as:
Horizontal density refers to the amount of backyard open space per m2 of developed area. It is measured through the coverage coefficient, i.e. the share of a land plot’s surface area occupied by the building footprint. It is calculated as:
Population density refers to the distribution of population per unit of area. Three different measures of population are used: population density per square kilometre of land area; population density per square kilometre of built area and population density per square kilometre of floor space.
Notes
← 1. Fixed individual preference parameters is a standard assumption in dynamic simulation or econometric models. It should be noted that the assumption of fixed preferences does not imply that any of the individual choices remains fixed over time. The latter evolve as economic constraints and incentives change over time. Stated differently, preferences refer to the technical parameters of the model that, along with other economic aspects (e.g. prices), affect the observable behaviour of individuals, i.e. the actual choices.
← 2. Historically, technical progress in improving the fuel economy of ICE vehicles has been partly offset by preferences for larger and more powerful cars. Over the past thirty years, however, there has been a clear trend towards improving improved vehicle fuel economy. This has coincided with policy initiatives, which set technical and efficiency standards for new cars but vary at a national level. Since the New Zealand vehicle fleet is composed of imported cars, the majority of which are produced in Japan, changes to the fuel economy are affected by changes in the fuel economy of Japanese cars.
← 3. The formula to calculate CO2e emissions from electric buses is: where R is the occupancy rate.
← 4. This is calculated by dividing the total passenger kilometres driven by bus in 2018 with the estimated vehicle kilometres generated by buses. The latter is approximated by the ratio of the total litres of diesel consumed by buses in 2018 to the fuel economy of a representative bus.
← 5. This is a limitation of the model, leading to an overestimation of emissions from public transport. That limitation should be considered in the evaluation of policies that increase public transport ridership in 2030 and 2050. However, the discrepancy is not sizeable, due to the assumption of public transport’s gradual electrification.
← 6. The benchmark year of the study is 2018. Where 2018 data is not available, data from the closest available year is used instead.
← 7. The VAT paid on the purchase price of a new car (NZD 4 682) is calculated by multiplying the pre-tax price of an average new car in 2018, estimated at NZD 31 210, with the VAT of 15%.
← 8. This is done by dividing the corresponding lump-sum fees by the assumed kilometric lifespan of a car (200 000 km).
← 9. Chapter 3 provides a full discussion of the calculation of the welfare outcomes of each of these policy packages.
← 10. For instance, the average annual wage in Sweden is approximately EUR 36 500, labour supply per worker is 1 609 hours per year, or approximately 201 days per year. That implies an average daily gross income of € 181.6. Assuming an average daily toll cost of EUR 2.9 (SEK 30) implies a daily cost of approximately 1.6% of gross income.
← 11. The actual fares paid for a trip taken with public transport in Auckland depend on the number of zones the passenger traverses during the trip. That number correlates, but does not coincide, with trip distance. The subsidy reduces the per-zone price relative to the BAU by 80%. It should not be confused with a fixed subsidy per kilometre.
← 12. For instance, densifying a 200 m2 land plot, which is 50% covered by a five-storey building with the widespread densification program produces 625 m2 of additional residential floor space. The corresponding increase for an identical land plot whose surface is covered by 30% by a two-storey building is 150 m2.
Metadata, Legal and Rights
https://doi.org/10.1787/095848a3-en
© OECD 2021
The use of this work, whether digital or print, is governed by the Terms and Conditions to be found at http://www.oecd.org/termsandconditions.